, Volume 27, Issue 4, pp 400-405,
Open Access This content is freely available online to anyone, anywhere at any time.
Date: 12 Feb 2013

FDG PET/CT evaluation of pathologically proven pulmonary lesions in an area of high endemic granulomatous disease

Abstract

Purpose

The goal of this study is to assess how reliable the threshold maximum standardized uptake value (maxSUV) of 2.5 on positron emission tomography–computed tomography (PET/CT) is for evaluation of solitary pulmonary lesions in an area of endemic granulomatous disease and to consider other imaging findings that may increase the accuracy of PET/CT.

Materials and methods

The staging PET/CT of 72 subjects with solitary pulmonary lesions (nodules (less than 3 cm) or masses (greater than 3 cm)) were retrospectively reviewed. Pathology proven diagnosis from tissue samples was used as the gold standard. Logistic regression was used to assess whether the subject’s age, maxSUV, size of lesion, presence of emphysema, or evidence of granulomatous disease was predictive of malignancy.

Results

Malignant lesions were identified in 84.7 % (61/72) of the 72 subjects. A threshold maxSUV of 2.5 had a sensitivity of 95.1 % (58/61), specificity of 45.5 % (5/11), positive predictive value of 90.6 % (58/64), negative predictive value of 62.5 % (5/8) and an accuracy of 87.5 % (63/72). The false negative rate was 4.9 %, and the false positive rate was 54.5 %. All 3 false negatives were less than or equal to 1.0 cm; however, false positives ranged from 1.1 to 5.6 cm. The false negatives had a mean (SD) maxSUV of 2.0 (0.4), whereas the false positives had a mean (SD) maxSUV of 5.6 (3.0). Emphysema was associated with 1.1 higher odds of malignancy, and evidence of granulomatous disease was associated with 0.34 lower odds of benign disease, however, neither was statistically significant (p = 0.92 and p = 0.31, respectively). Higher maxSUV was significantly associated with increased risk of malignancy (p = 8.3 × 10−3). Older age and larger size of lesion were borderline associated with increased risk of malignancy (p = 0.05 and p = 0.07, respectively).

Conclusion

In an area of high endemic granulomatous disease, the PET/CT threshold maxSUV of 2.5 retains a high sensitivity (95.1 %) and positive predictive value (90.6 %) for differentiating benign from malignant pulmonary lesions; however, the specificity (45.5) and negative predictive value (62.5) decrease due to increased false positives. The presence of emphysema and absence of evidence of granulomatous disease increases the probability that a pulmonary lesion is malignant; however, these were not statistically significant.